Laminin and fibronectin concentrations of the follicular fluid correlate with granulosa cell luteinization and oocyte quality

Abstract

Background and Aims:  Progesterone production of human cultured luteinizing granulosa cells was reported to be modified by extracellular matrix, suggesting that extracellular matrix regulates luteinization of granulosa cells after ovulation. In the present study, the relationship among laminin, fibronectin, progesterone and estradiol in follicular fluid along with oocyte quality was analyzed to estimate the physiological role of extracellular matrix in follicular luteinization and oocyte quality during ovulation.

Methods and Results:  Follicular fluid was collected at oocyte pick‐up from the patients undergoing in vitro fertilization treatment and intracytoplasmic sperm injection. The concentrations of laminin, fibronectin, progesterone and estradiol in the follicular fluid were measured by enzyme immunoassay and radioimmunoassay. The morphology of oocytes were also assessed during the procedure of intracytoplasmic sperm injection and was classified into normal and abnormal groups. The fibronectin concentration was higher in the normal ooplasm group than in the abnormal group, but it did not correlate with estradiol or progesterone concentration. However, laminin concentration significantly correlated with that of progesterone, but not with cytoplasm morphology of oocytes. There was no difference in estradiol or progesterone concentration between the normal and abnormal groups.

Conclusion:  These findings suggest that extracellular matrix plays some roles in regulating human granulosa cell luteinization and oocyte quality during ovulation. (Reprod Med Biol 2004; 3: 43–49)

INTRODUCTION

IMPLANTATION RATE OF embryos in human in vitro fertilization and embryo transfer (IVF‐ET) treatment has been reported to be approximately 15%, which is still lower than in normal fertile couples. The morphology of oocytes, as well as embryos, was proposed to be a critical factor for the success in IVF‐ET therapy ¹ , ² and much attention has been focused on the control of oocyte maturation. However, there are few effective methods to clinically estimate the quality of oocytes except for morphological assessment. The quality of oocytes is considered to be deeply related to and is affected by the surrounding circumstances produced by granulosa and theca cells. It is therefore important to elucidate the functions of these cells and their relationship with oocyte quality especially in periovulatory phase, when the oocytes dramatically change their morphology and function.

It is generally accepted that the differentiation of granulosa and theca cells is mainly regulated by gonadotropins, and also by growth factors and cytokines. ³ To investigate new regulatory factors involved in ovarian cell differentiation, we raised monoclonal antibodies against human granulosa cells. By analyzing the antigens that were detected by the monoclonal antibodies, it was revealed that integrin α6β1 was specifically expressed on human granulosa cells in preovulatory follicles and corpora lutea of the early luteal phase. ⁴ , ⁵ Later laminin, the ligand for integrin α6β1, was also shown to be expressed on granulosa cells in the periovulatory phase. The production of progesterone by cultured granulosa cells was suppressed by laminin via interaction with integrin α6β1. ⁶ In murine ovary, integrin α6 was expressed on granulosa cells of the primordial, primary and secondary follicles. The administration of anti‐integrin α6 antibody, which inhibits the interaction between integrin α6 and laminin, promoted the follicular growth in response to gonadotropins. ⁷ In addition, integrin α5β1 and its ligand fibronectin were also shown to appear rapidly on human granulosa cells during the ovulation. Their expressions continued during the early luteal phase, then gradually decreased in the midluteal phase. The expression of integrin α5 on granulosa cells was enhanced by human chorionic gonadotropin (hCG) in vitro, suggesting that its expression is induced by luteinizing hormone surge in vivo. ⁸ These findings indicated that granulosa cells express receptors for laminin and fibronectin and suggested that the interaction with these extracellular matrices (ECM) is involved in the regulation of granulosa cell differentiation. ⁹ , ¹⁰ Human oocytes are also reported to express receptors for laminin and fibronectin. ¹¹

In the current study, to estimate the physiological roles of ECM on oocytes and granulosa cells in the periovulatory phase, we examined correlation between follicular fluid concentrations of laminin and fibronectin and the morphology of oocyte or follicular concentrations of steroid hormones.

MATERIALS AND METHODS

Retrieval of oocytes and follicular fluid

OOCYTES AND FOLLICULAR fluid were retrieved from the patients who underwent intracytoplasmic sperm injection (ICSI) at Kyoto University. Ovulation induction was performed as described previously. ¹² Briefly, patients receiving a gonadotropin releasing hormone analog (buserelin acetate, Aventis Pharma, Tokyo, Japan) from the first day of the cycle, were hyperstimulated with pure follicle‐stimulating hormone (FSH) (Serono, Tokyo, Japan) or human menopausal gonadotropin (Organon, Tokyo, Japan) until the follicles reached maturity. Follicles were aspirated 36 h after the administration of hCG (Mochida Pharmaceutical, Osaka, Japan). The samples of follicular fluid were obtained from one or two follicle(s) per patient. To avoid the contamination of the other follicular fluid, only the follicular fluids that were initially punctured in the unilateral side were used for the present study. The sample of the follicular fluid in the contralateral ovary was collected after changing a disporsal pick‐up tube unit. The collected samples were immediately centrifuged at 1000 g for 10 min, and the supernatant was stored at −80°C until the assay. Informed consent was obtained from all patients prior to the study. Analysis of these samples was approved by the Ethical Committee of Kyoto University Hospital.

Morphology of the oocytes

Two hours after the oocyte retrieval, cumulus‐oocyte complex was treated with 0.05% hyaluronidase, and the cumulus‐removed oocytes were observed under light microscopy (Nikon, Tokyo, Japan) at the time of ICSI. Only mature oocytes (metaphase II) were included in the current study, and immature oocytes or empty follicles were excluded. The morphology of cytoplasm of oocytes were assessed as reported by Veeck. ² , ¹³ Cytoplasm with excessive granularity, with large perivitelline space, with central darkness, or with cytoplasmic vacuole was classified as abnormal in morphology. Cytoplasm without these findings was classified as normal in morphology.

Follicular fluid concentrations of laminin, fibronectin, estradiol and progesterone

Follicular fluid concentrations of laminin and fibronectin were measured by enzyme immunoassay kits (Fuji Chemical Industries, Toyama, Japan and Paesel Lorei, Hanau, Germany, respectively). Inter‐ and intra‐assay coefficients of variation were 8.5% and 6.2% for the laminin assay and 5.6% and 4.5% for the fibronectin assay, respectively. The concentrations of estradiol and progesterone were measured using radioimmunoassay kits (Daiichi Radio Isotope Research, Tokyo, Japan). Inter‐ and intra‐assay coefficients of variation were 6.5% and 5.3% for the progesterone assay and 7.4% and 6.3% for the estradiol assay, respectively.

Statistics

The data were expressed as means ± SD, and were analyzed by a two‐tailed unpaired t‐test and by a simple regression. Differences were regarded as significant at the 5% level.

RESULTS

FORTY‐SEVEN MATURE oocytes at metaphase II were analyzed. They were retrieved from 29 patients, whose age ranged from 24 to 41 years, with the mean age of 35.3 ± 3.0 years. The volume of follicular fluids ranged from 2.5 to 8.0 mL, with a mean of 4.2 ± 1.5 mL. The cytoplasm morphology of 47 oocytes were classified as normal (n = 30) and abnormal (n = 17). Abnormal findings were excessive granularity (n = 7), central darkness (n = 2), vacuole (n = 1), or large perivitelline space (n = 7) in the cytoplasm. The quality of ooplasm was not significantly correlated with age or with follicular fluid volume (35.6 ± 2.7 years vs 34.7 ± 3.7 years or 4.3 ± 1.6 mL vs 3.9 ± 1.2 mL in the normal morphology group and in the abnormal group, respectively).

The ECM levels in follicular fluid were compared between the groups of oocytes with normal and abnormal cytoplasmic morphology. In the group of oocytes with normal cytoplasm, follicular fluid concentration of fibronectin was higher than in that with abnormal cytoplasm (P < 0.01, Fig. 1a). However, there was no difference in laminin concentration between the normal ooplasm group and the abnormal group (Fig. 1b).

Figure 1.

The relationship between the cytoplasm morphology of the oocytes observed at the time of intracytoplasmic sperm injection and the follicular fluid concentrations of extracellular matrices. (a) Fibronectin level was significantly higher in the normal ooplasm morphology group than in the abnormal group (P < 0.01). In contrast, (b) laminin concentration was not significantly different (N.S.) between the two groups.

Steroid hormone levels in follicular fluid were analyzed between the groups of oocytes with normal and abnormal cytoplasmic morphology. There was no difference in follicular fluid progesterone concentration between the normal morphology group and the abnormal group (Fig. 2a). Follicular fluid estradiol concentration appeared to be higher in the normal morphology group than in the abnormal group, but it was not statistically significant (P = 0.062, Fig. 2b). There was no difference in progesterone/estradiol or estradiol/progesterone ratio between the two groups.

Figure 2.

The relationship between the cytoplasm morphology of the oocytes observed at the time of intracytoplasmic sperm injection and the follicular fluid concentrations of steroid hormones. (a) Progesterone and (b) estradiol levels were not significantly different (N.S.) between the normal ooplasm morphology group and the abnormal group.

The relation between follicular fluid ECM level and follicular fluid steroid hormone level were examined. There was no correlation between fibronectin and estradiol, or between fibronectin and progesterone (data not shown). However, there was a significant correlation between follicular fluid laminin level and progesterone level (n = 47, R ² = 0.141, P < 0.01, Fig. 3), whereas there was no correlation between laminin and estradiol (data not shown).

Figure 3.

The relationship between the follicular fluid concentrations of laminin and progesterone. Laminin concentration was significantly correlated with that of progesterone in the follicular fluid (P < 0.01).

DISCUSSION

THE CURRENT STUDY showed that the morphology of cytoplasm of oocyte was closely related with the follicular fluid concentration of fibronectin. It is reported that fibronectin is produced by granulosa cells in some animals, and that the production increases or decreases during follicular development under the regulation by gonadotropin. ¹⁴ , ¹⁵ , ¹⁶ In humans, Lobb and Dorrington reported that fibronectin was detected in the culture supernatant of granulosa cells obtained from growing follicles more than 10 mm in diameter, and that the fibronectin production is influenced by (Bu)2 cyclic adenosine monophosphate, suggesting that human granulosa cells produce fibronectin and that this production changes during follicular development. ¹⁷ We previously reported that immunohistochemical detection of fibronectin was rapidly increased among luteinizing granulosa cells during ovulation. ⁸ Familiari et al. showed by immunohistochemistry that fibronectin was expressed on the cell surface and in the cytoplasm of the cumulus cells obtained from preovulatory follicles of IVF‐undergoing patients. ¹⁸ Tsuiki et al. investigated the correlation between follicular maturity and follicular fluid concentration of fibronectin in IVF‐treated patients. ¹⁹ They classified mature and immature follicles, as having loose and tight cumulus cells, respectively, and showed that follicular fibronectin level is higher in mature follicles. These findings suggest that fibronectin has some physiological roles on the dramatic events induced by LH surge. Fibronectin receptor was shown to be expressed on human oocytes. ¹¹ The immunohistochemical study also demonstrated that fibronectin exists on the cell surface of oocyte as well as on cumulus cells. ¹⁸ We confirmed that integrin α5β1, a fibronectin receptor, was expressed on human luteinizing granulosa cells of preovulatory follicles and its expression was augmented by hCG. ⁸ In the present study, although the morphology of ooplasm showed no relation with follicular fluid volume, estradiol or progesterone level, it was closely related with the fibronectin level. Consequently, it can be proposed that follicular fluid fibronectin level is a new parameter to estimate oocyte quality, which is independent of steroid hormone levels. The mechanism how fibronectin is associated with ooplasm morphology is still unclear. Considering the distribution of fibronectin and its receptor on oocyte and granulosa cells including the cumulus cells, fibronectin may have an effect on ooplasm morphology, either indirectly via regulation of granulosa cell differentiation or directly by acting on oocyte.

The current study also showed that follicular fluid concentration of progesterone is positively related with that of laminin. This is compatible with our previous observation that the expression of laminin among luteinizing granulosa cells rapidly increased during ovulation process. ⁶ Previously, laminin was shown to suppress progesterone production by luteinizing granulosa cells via interaction with integrin α6β1, implying the involvement of laminin in the regulation of granulosa cell luteinization. ⁶ After LH surge, granulosa cell luteinization is started, and the rise of progesterone production was shown to be important as a trigger for the following inflammatory cascade of follicular rupture. ²⁰ , ²¹ However, the factors responsible for resumption of oocyte meiosis is still unclear, although some changes in granulosa cells induced by LH surge are considered essential for it. ²² The processes of oocyte meiosis, granulosa cell differentiation, and follicular rupture should be timely coordinated to achieve fertile conditions where the ovulated oocyte in the pelvic cavity is a stage of metaphase II and is surrounded by expanding cumulus cells rich in proteoglycans. ²³ , ²⁴ Thus, the regulation of granulosa cell differentiation during LH surge must be an important issue. The only established promoting factor for granulosa cell luteinization is LH/hCG. The speed of luteinization of granulosa cells should be bidirectionally regulated to orchestrate other processes of oocyte meiosis and follicular rupture. Laminin produced by luteinizing granulosa cells may be candidates for the negative feedback regulation. As laminin suppresses progesterone production by luteinizing granulosa cells in vitro, we speculate that follicular fluid concentration of laminin increases during luteinization of granulosa cells in preovulatory follicles, and that the accelerated differentiation of granulosa cells and the increase of progesterone concentration is negatively regulated by laminin.

In conclusion, we showed that the follicular fluid concentration of fibronectin correlated with the morphology of oocyte cytoplasm. In addition, the follicular fluid concentration of laminin correlated with that of progesterone. These findings support the physiological significance of ECM in granulosa cell luteinization and oocyte quality.